Fear learning permits an animal to respond adaptively to threatening circumstances (Bolles 1970). However, dysregulation of the brain's systems for fear learning can lead to debilitating anxiety disorders (Rosen and Schulkin 1998), including post-traumatic stress disorder (PTSD). In preliminary experiments in a mouse model of olfactory fear learning, longitudinal optical imaging experiments revealed large, associative, stimulus- specific increases in neurotransmitter release from olfactory sensory neurons (OSNs) when they are stimulated by fear-associated odors (CS+) in vivo. This neural response to the CS+ is enhanced after conditioning both compared to responses to neutral odors and compared to its own pre-conditioning baseline, thus showing that fear conditioning selectively changes the neural representation of the CS+ at the input to the brain. After fear conditioning, these primary sensory responses to footshock-predictive odors were actually larger than could be evoked by any concentration of that odor under control circumstances, perhaps serving as a warning signal to enhance reaction to the CS+ or to draw attention to it. This result suggests that changes in sensory processing of threat-related stimuli could play a role in anxiety disorders, which include a ubiquitous attentional bias toward dangerous or unpleasant stimuli. This project will combine optical neuroimaging, behavioral, and pharmacological techniques to investigate this learning-induced sensory neuroplasticity. The Specific Aims of the project are 1) to use various fear conditioning paradigms to determine the circumstances under which this sensory neuroplasticity occurs and how it relates to the formation of memories relating the odor and fear-inducing stimuli, 2) to learn the neural mechanisms by which this sensory neuroplasticity occurs, including the circuitry by which information about the fearful associations of CS+ reaches the primary sensory neurons, and 3) to discern how the change in the neural representation of the CS+ alters the brain's representation of sensory input both after normal emotional learning and in a rodent model of PTSD.